Use of lectin binding characteristics to identify gastrointestinal parasite eggs in faeces

Rapid, automated quantification of Haemonchus contortus ova in sheep faecal samples

Haemonchus contortus is one of the most pathogenic nematodes affecting small ruminants globally and is responsible for large economic losses in the sheep and goat industry. Anthelmintic resistance is rampant in this parasite and thus parasite control programs must account for drug efficacy on individual farms and, sometimes, whether H. contortus is the most prevalent trichostrongylid. Historically, coproculture has been the main way to determine the prevalence of H. contortus in faecal samples due to the inability to morphologically differentiate between trichostrongylid egg types, but this process requires a skilled technician and takes multiple days to complete. Fluoresceinated peanut agglutinin (PNA) has been shown to specifically bind H. contortus and thus differentiate eggs based on whether they fluoresce, but this method has not been widely adopted. The ParasightTM System (PS) fluorescently stains helminth eggs in order to identify and quantify them, and the H. contortus PNA staining method was therefore adapted to this platform using methodology requiring only 20 min to obtain results. In this study, 74 fecal samples were collected from sheep and analyzed for PNA-stained H. contortus, using both PS and manual fluorescence microscopy. The percentage of H. contortus was determined based on standard total strongylid counts with PS or brightfield microscopy. Additionally, 15 samples were processed for coproculture with larval identification, and analyzed with both manual and automated PNA methods. All methods were compared using the coefficient of determination (R2) and the Lin's concordance correlation coefficient (ρc). ParasightTM and manual PNA percent H. contortus results were highly correlated with R2 = 0.8436 and ρc = 0.9100 for all 74 fecal samples. Coproculture versus PS percent H. contortus were also highly correlated with R2 = 0.8245 and ρc = 0.8605. Overall, this system provides a rapid and convenient method for determining the percentage of H. contortus in sheep and goat fecal samples without requiring specialized training.

Combined Molecular and Lectin Binding Assays to Identify Different Trichostrongyle Eggs in Feces of Sheep and Goats from Egypt

BACKGROUND

Trichostrongyles are common causes of parasitic gastroenteritis in sheep and goats worldwide. Accurate identification of these nematodes to the genus and/or species level is important for therapy selection and control strategies. In the present study, molecular and egg-lectin binding approaches were employed to identify the most economically important trichostrongyles circulating in sheep and goat herds from six districts in Dakahlia governorate, Egypt.

MATERIALS

Fecal samples from 653 and 205 goats reared within 17 herds were collected and tested for the trichostrongyle eggs using the modified Wisconsin sucrose flotation method. For identification of the trichostrongyle(s) present, eggs from 75 (63 sheep and 12 goats) samples which had high egg count (EPG) and pooled eggs (n = 19 pools, 15 sheep and 4 goats) from samples with moderate or low EPGs were examined. Molecular examination was conducted amplifying the ITS2 region of the rDNA for six different trichostrongyles in individual PCR reactions. For egg-lectin bindings, 4 fluorescently-labeled specific lectins were used; peanut agglutinin (PNA) for Haemonchus contortus, Aleuria aurantia agglutinin (AAL) for Trichostrongylus species, Lens culinaris agglutinin (LCA) for Teladorsagia circumcnicta and Lotus tetragonolobus lectin (LTL) for Cooperia species.

RESULTS

Fourteen (82.3%) herds were found infected, of which trichostrongyle eggs were detected in fecal samples of 26.5% (173/653) of sheep and 10.2% (21/205) of goats. Results of the PCR and lectin bindings were compatible and 4 trichostrongyles were detected: H. contortus, T. circumcincta, Trichostrongylus axei and Trichostrongylus colubriformis. Haemonchus contortus eggs were found in all the infected herds, and as the single species in 21 and 5 of sheep and goat samples, respectively. Lectin stained smears demonstrated the dominance of H. contortus eggs over eggs of the other detected trichostrongyles. Eleven herds were found infected with T. axei as the second most prevalent trichostrongyle; however, few AAL-stained eggs were noticed in the positive samples. Mixed infections were frequently detected as H. contortus-T. axei combination. Infections with T. circumcincta were noted in sheep samples from two herds, but not in any sample from the goats. No Ostertagia leptospicularis, Cooperia curticei or Nematodirus species were noted among the tested samples.

CONCLUSIONS

This is the first molecular and lectin binding survey to determine the species composition of trichostrongyles infecting sheep and goats from Egypt. Haemonchus contortus plays the principal role in small ruminant trichostrongylosis in Egypt. Egg-lectin staining shows promise for future for its application in routine diagnosis as a rapid and simple technique. Findings of the earlier reports from Egypt are tabulated and reviewed.

Triple lectin staining of trichostrongyle eggs from naturally infected small ruminants

Trichostrongyle nematodes can be a major threat to the profitability of small ruminant producers depending of the species and intensity of trichostrongyles parasitizing their herd. Haemonchus contortus, Teladorsagia circumcincta, and Trichostrongylus colubriformis are typically the most common and clinically important species. Three lectins (PNA, LCA and AAL) have been reported to bind specifically to eggs from these three genera and therefore could be used to quantify the intensity of each species in individual animals. Peanut agglutinin (PNA) has been the most commonly tested lectin because it selectively binds intensely to eggs of the most pathogenic species, H. contortus. Lens culinaris agglutinin (LCA) and Aleuria aurantia agglutinin (AAL) have shown specificity to T. circumcincta and Trichostrongylus spp. respectively, however, these lectins have only been evaluated using eggs harvested directly from adult females, and not from fecal samples. The purpose of the present study is to describe a method to sequentially stain H. contortus, T. circumcincta and Trichostrongylus spp. fecal eggs with PNA, LCA and AAL, and then evaluate the resultant staining patterns seen with eggs collected from a naturally infected goat shown with PCR to contain H. contortus, T. circumcincta, Ostertagia leptospicularis, Trichostrongylus colubriformis and Trichostrongylus axei eggs. These results were also compared with patterns observed with eggs stained with single lectins and double combinations of lectins. The various patterns were then compared to those seen with egg samples collected from an ewe shown to only contain H. contortus. PNA bound intensely and uniformly to all eggs from samples containing only H. contortus eggs; however, some eggs additionally bound LCA and AAL in localized patches of varying size, and a few eggs exhibited intense and uniform binding of all three lectins. Single PNA-staining of goat samples containing the five trichostrongyles species identified most eggs as H. contortus, and triple-staining showed patterns consistent with those seen for H. contortus. Binding of AAL to non-Haemonchus eggs was uniform but showed significant variations in intensity. Lesser staining eggs tended to also stain intensely with LCA, which is consistent with published binding pattern for T. circumcincta. Most eggs that AAL bound intensely to did not bind with LCA, which is consistent with published binding pattern for Trichostrongylus spp. Autofluorescence was observed with the DAPI filter-cube among most non-Haemonchus eggs. This study demonstrates the need for additional field studies to further validate the specificity of these three lectins for use in identifying eggs from the three species of trichostrongyles.

Diarrhoea associated with gastrointestinal parasites in grazing sheep

Diarrhoea is a common, widespread and frustrating reality for sheep enterprises in most sheep producing regions globally and of particular concern in Australia as the major risk factor for breech flystrike. Parasitic disease has long been recognised as an important factor in diarrhoea in sheep, particularly the gastrointestinal nematodes (Trichostrongylus and Teladorsagia species). This review focuses on the role of parasitic infections in causing diarrhoea in sheep, with emphasis on the epidemiology of diarrhoea outbreaks related to worms and opportunities to manage the risk of diarrhoea outbreaks in sheep related to parasitic infections. Parasitic nematodes damage the gastrointestinal tract via a complex relationship between direct impacts from worms, such as physical changes to the gut mucosa, and indirect effects largely associated with the host response. Diarrhoea associated with large worm burdens is most efficiently managed through integrated parasite management programs. Despite some limitations, measuring faecal worm egg counts remains a mainstay for assessing the contribution of worms to outbreaks of diarrhoea in sheep. Larval hypersensitivity scouring is emerging as a significant cause of worm-related diarrhoea in sheep without large adult worm burdens in some geographic locations. The syndrome describes a heightened inflammatory response to the ingestion of trichostrongylid infective larvae seen in the gut of sheep with diarrhoea, and is most effectively addressed through selecting sheep for low breech soiling ('dag scores'), as worm resistant sheep may show an increased propensity for diarrhoea, even with low rates of larval challenge. Importantly, dag should be considered as a separate trait to WEC in breeding indexes. Outbreaks of diarrhoea in young sheep are often multifactorial, and co-infections with nematodes and other infectious agents associated with diarrhoea are common. This presents challenges for the field investigation of diarrhoea in grazing sheep.

Egg autofluorescence and options for detecting peanut agglutinin binding for the identification of Haemonchus contortus eggs in fecal samples

Quantifying eggs from Haemonchus and other trichostrongyle genera in sheep and goat fecal samples is important for evaluating control and treatment strategies for this family of nematodes with divergent pathologies, capabilities for anthelmintic resistance and environmental susceptibilities. Unfortunately, egg morphology among most of the genera do not differ enough to support the accurate identification of these genera with standard microscopic techniques. Several studies have identified specific lectins which bind selectively to sugars located on the egg surfaces for individual genera among the trichostrongyles. To detect lectins binding to these eggs, they must be directly or indirectly bound to fluorophores, and observed with an epi-fluorescence microscope. The binding of multiple lectins to isolated eggs from a fecal sample can be simultaneously detected if fluorophores are used whose excitation and emission spectra do not overlap, and this would enable the development of a fluorescence-based diagnostic test that identifies multiple trichostrongyle genera within each sample. The present study compared the usefulness of different, commercially available detection systems for use in detecting lectin binding to trichostrongyle eggs. Comparisons were made using the detection of PNA binding to H. contortus eggs with the goal of finding three systems with color spectra that do not overlap. These evaluations included both fluorophores directly conjugated to PNA in a one-step incubation protocol and a two-step incubation protocol involving biotinylated PNA and streptavidin conjugated to different fluorophores. Autofluorescence can affect the efficiency of any fluorescence-based detection system, and significant autofluorescence was observed among the unstained H. contortus eggs with the DAPI-type fluorescence filter, but it was significantly lower with the FITC-type filter and was virtually absent with the rhodamine-type filter. This study demonstrated that all the PNA detection methods tested with H. contortus eggs generated fluorescence intensities (FIs) that were significantly above the autofluorescence generated by the eggs among the three different fluorescence filters. Fluorescence intensities from PNA directly conjugated to either the FITC or rhodamine fluorophores were not different, but the lower autofluoresence in the rhodamine-type filter will enable this fluorophore to be detected more efficiently. Use of biotinylated PNA combined with streptavidin-conjugated to synthetic fluorophores (Alexa Fluor 405, 488 and 546) significantly increased FIs over that of the directly conjugated PNA, but there were no significant differences in FIs among these three biotin-avidin conjugation fluorophores. This biotin-avidin system required two incubation steps. Doubling the concentration of PNA also provided increased FI, at least for the biotin-avidin system. Adding an additional amplification step to the biotin-avidin system involving biotinylated anti-streptavidin followed by the streptavidin-Alexa Fluor complex also provided additional fluorescence.

High-throughput identification and quantification of Haemonchus contortus in fecal samples

Haemonchus contortus are gastrointestinal nematodes of the family Trichostrongylidae that naturally infect small ruminants while grazing, posing a risk to both animal health and farm profitability. Current diagnostics depend on exacting lab techniques, including manual egg counts and larval differentiation, all of which require time, effort, and specialized technicians. The goal of this study was to facilitate and accelerate the identification and quantification of H. contortus in fecal samples through the use of fluorescein-isothiocyanate peanut-agglutinin staining in order to allow automated detection using a 96-well microplate reader. Next, the model was to be validated using samples containing unknown quantities of eggs. Automated analysis of fluorescence emission of known quantities of H. contortus eggs confirmed an almost perfect linear correlation (r = 0.9984, p < 0.0001), indicating that this new approach can satisfactorily be used to quantify H. contortus eggs on a comparative fluorescence scale. As validation, clinical samples containing an unknown quantity of H. contortus eggs were then analyzed by comparing two methods: either Wisconsin Sugar Flotation (WSF) and McMaster counting followed by manual fluorescence microscopy, or WSF coupled with automated microplate reading. Pearson analysis revealed highly significant correlation between manual and automated methods (r = 0.9999, p < 0.0001), while Bland-Altman plots demonstrated excellent agreement between the two (bias = -0.817 ± 9.94 with 95% limits of agreement from -20.31 to 18.67). Overall, these results demonstrate that high-throughput screening fluorescence detection and quantification of H. contortus eggs is both accurate and rapid.

Comparison of Four Diagnostic Methods for Detection and Relative Quantification of Haemonchus contortus Eggs in Feces Samples

We compared four methods for identification of Haemonchus contortus eggs. With increased trade in animals within and between countries and continents, it has become important to correctly identify H. contortus eggs in fecal samples. To validate the outcome of diagnostic tests, sheep feces (n = 38) were collected from naturally infected flocks in Sweden. Subsamples were analyzed with (a) McMaster egg counting; (b) differential counting of eggs after staining with peanut agglutinin (PNA); (c) detection of DNA following amplification by real-time quantitative polymerase chain reaction (qPCR); and (d) loop-mediated isothermal amplification (LAMP). Differences between similar tests (microscopic and molecular) and SD (±SD) were analyzed with Bland-Altman plots and Spearman rank correlation. Strongylid egg counts ranged from 200 to 12,100 eggs per gram (epg) (mean epg ± SD = 1,278 ± 2,049). Microscopy showed presence of H. contortus eggs in 27 (73%) unstained samples and in 28 (76%) samples stained with PNA, whereas 29 samples (78%) tested positive in LAMP and 34 (91%) in qPCR analysis. The cycle threshold (Ct) values with LAMP ranged between 13 and 38 (mean ± SD = 21 ± 7), and those in qPCR between 25 and 49 (mean ± SD = 33 ± 6). In the LAMP and qPCR analyses, seven (19%) and three (8%) samples, respectively, had a cycle threshold (Ct) >35, whereas no reactions were observed in eight (22%) and three (8%) samples, respectively. There was good agreement between the diagnostic tests based on microscopic examination and DNA detection, although the molecular tests were more sensitive. The bias between the microscopy methods (-4.2 ± 11) was smaller than for the molecular tests (-9.8 ± 10). The observed ranking in terms of test sensitivity was: McMaster counting by conventional microscopy < PNA < LAMP < qPCR. In conclusion, H. contortus can be identified by McMaster counting, without major mistakes regarding false positive results. However, molecular methods provide the capacity to diagnose H. contortus eggs with increased accuracy. This is essential when animals are investigated in quarantine or in studies evaluating anthelmintic treatment efficacy. These methods could also be applied to fecal samples from wildlife to investigate nematode transmission between wildlife and livestock.

Diagnosis, Treatment and Management of Haemonchus contortus in Small Ruminants

Haemonchus contortus is a highly pathogenic, blood-feeding nematode of small ruminants, and a significant cause of mortalities worldwide. Haemonchosis is a particularly significant threat in tropical, subtropical and warm temperate regions, where warm and moist conditions favour the free-living stages, but periodic outbreaks occur more widely during periods of transient environmental favourability. The clinical diagnosis of haemonchosis is based mostly on the detection of anaemia in association with a characteristic epidemiological picture, and confirmed at postmortem by the finding of large numbers of H. contortus in the abomasum. The detection of impending haemonchosis relies chiefly on periodic monitoring for anaemia, including through the 'FAMACHA' conjunctival-colour index, or through faecal worm egg counts and other laboratory procedures. A range of anthelmintics for use against H. contortus is available, but in most endemic situations anthelmintic resistance significantly limits the available treatment options. Effective preventative programmes vary depending on environments and enterprise types, and according to the scale of the haemonchosis risk and the local epidemiology of infections, but should aim to prevent disease outbreaks while maintaining anthelmintic efficacy. Appropriate strategies include animal management programmes to avoid excessive H. contortus challenge, genetic and nutritional approaches to enhance resistance and resilience to infection, and the monitoring of H. contortus infection on an individual animal or flock basis. Specific strategies to manage anthelmintic resistance centre on the appropriate use of effective anthelmintics, and refugia-based treatment schedules. Alternative approaches, such as biological control, may also prove useful, and vaccination against H. contortus appears to have significant potential in control programmes.

Development of a loop-mediated isothermal amplification (LAMP) assay for the sensitive detection of Haemonchus contortus eggs in ovine faecal samples

A major constraint on the effective control and management of helminth parasites in livestock is the lack of rapid and reliable diagnostic tests to identify the parasite species responsible for disease and to allow informed treatment decisions to be made. In the present study, we have developed a novel DNA-based assay for the detection of Haemonchus contortus eggs in ovine faecal samples, using loop-mediated isothermal amplification (or LAMP). LAMP allows for rapid detection of H. contortus DNA under isothermal incubation conditions. The robust nature of this assay negates the need for extensive DNA extraction, allowing amplification from relatively crude samples. Preliminary results suggest that LAMP is highly specific, and does not cross-react with DNA from other common co-infecting parasites. The Haemonchus LAMP assay is also highly sensitive, exhibiting a 10 times lower detection limit than the equivalent PCR; 10(-5) ng/μl and 10(-4) ng/μl DNA, respectively, allowing detection in a faecal samples containing two Haemonchus eggs per gram. Translation of this assay onto a real-time platform provided rapid results and highlighted its potential as a quantitative assay which could inform treatment decisions in the future.

The Identification of Haemonchus Species and Diagnosis of Haemonchosis

Diagnosis is often equated with identification or detection when discussing parasitic diseases. Unfortunately, these are not necessarily mutually exclusive activities; diseases and infections are generally diagnosed and organisms are identified. Diagnosis is commonly predicated upon some clinical signs; in an effort to determine the causative agent, identification of genera and species is subsequently performed. Both identification and diagnosis play critical roles in managing an infection, and involve the interplay of direct and indirect methods of detection, particularly in light of the complex and expanding problem of drug-resistance in parasites. Accurate and authoritative identification that is cost- and time-effective, based on structural and molecular attributes of specimens, provides a foundation for defining parasite diversity and changing patterns of geographical distribution, host association and emergence of disease. Most techniques developed thus far have been grounded in assumptions based on strict host associations between Haemonchus contortus and small ruminants, that is, sheep and goats, and between Haemonchus placei and bovids. Current research and increasing empirical evidence of natural infections in the field demonstrates that this assumption misrepresents the host associations for these species of Haemonchus. Furthermore, the capacity of H. contortus to utilize a considerably broad spectrum of ungulate hosts is reflected in our understanding of the role of anthropogenic forcing, the 'breakdown' of ecological isolation, global introduction and host switching as determinants of distribution. Nuanced insights about distribution, host association and epidemiology have emerged over the past 30years, coincidently with the development of increasingly robust means for parasite identification. In this review and for the sake of argument, we would like to delineate the diagnosis of haemonchosis from the identification of the specific pathogen. As a foundation for exploring host and parasite biology, we will examine the evolution of methods for distinguishing H. contortus from other common gastrointestinal nematodes of agriculturally significant and free-ranging wild ruminants using morphological, molecular and/or immunological methods for studies at the species and genus levels.

The specific diagnosis of gastrointestinal nematode infections in livestock: larval culture technique, its limitations and alternative DNA-based approaches

The specific diagnosis of gastrointestinal nematode infections in ruminants is routinely based on larval culture technique and on the morphological identification of developed third-stage larvae. However, research on the ecology and developmental requirements of different species suggests that environmental conditions (e.g., temperature and humidity) for optimal development to occur vary between the different species. Thus, employing a common culture protocol for all species will favour the development of certain species over others and can cause a biased result in particular when species proportions in a mixed infection are to be determined. Furthermore, the morphological identification of L3 larvae is complicated by a lack of distinctive, obvious features that would allow the identification of all key species. In the present paper we review in detail the potential limitations of larval culture technique and morphological identification and provide account to some modern molecular alternatives to the specific diagnosis of gastrointestinal nematode infection in ruminants.

Current status for gastrointestinal nematode diagnosis in small ruminants: where are we and where are we going?

Gastrointestinal nematode (GIN) parasites pose a significant economic burden particularly in small ruminant production systems. Anthelmintic resistance is a serious concern to the effective control of GIN parasites and has fuelled the focus to design and promote sustainable control of practices of parasite control. Many facets of sustainable GIN parasite control programs rely on the ability to diagnose infection both qualitatively and quantitatively. Diagnostics are required to determine anthelmintic efficacies, for targeted treatment programs and selection of animals for parasite resistant breeding. This review describes much of the research investigated to date to improve the current diagnostic for the above practices which is based on counting the number of parasite eggs in faeces.

Oral dosing with papaya latex is an effective anthelmintic treatment for sheep infected with Haemonchus contortus

BACKGROUND

The cysteine proteinases in papaya latex have been shown to have potent anthelmintic properties in monogastric hosts such as rodents, pigs and humans, but this has not been demonstrated in ruminants.

METHODS

In two experiments, sheep were infected concurrently with 5,000 infective larvae of Haemonchus contortus and 10,000 infective larvae of Trichostrongylus colubriformis and were then treated with the supernatant from a suspension of papaya latex from day 28 to day 32 post-infection. Faecal egg counts were monitored from a week before treatment until the end of the experiment and worm burdens were assessed on day 35 post-infection.

RESULTS

We found that the soluble fraction of papaya latex had a potent in vivo effect on the abomasal nematode H. contortus, but not on the small intestinal nematode T. colubriformis. This effect was dose-dependent and at tolerated levels of gavage with papaya latex (117 μmol of active papaya latex supernatant for 4 days), the H. contortus worm burdens were reduced by 98%. Repeated treatment, daily for 4 days, was more effective than a single dose, but efficacy was not enhanced by concurrent treatment with the antacid cimetidine.

CONCLUSIONS

Our results provide support for the idea that cysteine proteinases derived from papaya latex may be developed into novel anthelmintics for the treatment of lumenal stages of gastro-intestinal nematode infections in sheep, particularly those parasitizing the abomasum.

Detection of Schistosoma mansoni eggs in feces through their interaction with paramagnetic beads in a magnetic field

Background

Diagnosis of intestinal schistosomiasis in low endemic areas is a problem because often control measures have reduced egg burdens in feces to below the detection limits of classical coproparasitological methods. Evaluation of molecular methods is hindered by the absence of an established standard with maximum sensitivity and specificity. One strategy to optimize method performance, where eggs are rare events, is to examine large amounts of feces. A novel diagnostic method for isolation of Schistosoma mansoni eggs in feces, and an initial evaluation of its performance is reported here.

Methodology/Principal Findings

Known amounts of S. mansoni eggs were seeded into 30 g of normal human feces and subjected to a sequence of spontaneous sedimentation, sieving, Ritchie method, incubation and isolation through interaction with paramagnetic beads. Preliminary tests demonstrated the efficacy of lectins as ligands, but they also indicated that the paramagnetic beads alone were sufficient to isolate the eggs under a magnetic field through an unknown mechanism. Eggs were identified by microscopic inspection, with a sensitivity of 100% at 1.3 eggs per gram of feces (epg). Sensitivity gradually decreased to 25% at a concentration of 0.1 epg. In a preliminary application of the new method to the investigation of a recently established focus in southern Brazil, approximately 3 times more eggs were detected than with the thick-smear Kato-Katz method.

Conclusions/Significance

The novel S. mansoni detection method may significantly improve diagnosis of infections with low burdens in areas of recent introduction of the parasite, areas under successful control of transmission, or in infected travelers. It may also improve the evaluation of new treatments and vaccines.

Schistosomiasis mansoni is a parasitic infection that affects approximately 200 million people, mainly in the tropics. The worms live inside the veins of intestines and liver and produce eggs that are eliminated within feces. If the eggs reach water, a ciliated larva is released and enters snails to develop into a larva infective to man and other vertebrates. Most infections evolve without overt disease, but severe intestinal, hepatic, pulmonary and cerebro-medulary dysfunctions may occur after many years. Definitive diagnosis is made through the identification of eggs in stool. Classical diagnostic methods fail to detect infection when the number of eggs is low (e.g., in areas where control measures have decreased the intensity of infection or in the case of light infections in travelers who have had only brief exposure). A new and very sensitive method is reported here, in which eggs are isolated from large amounts of feces through their interaction with magnetic beads. After incubation with the fecal sediment, eggs co-migrate with the beads towards a magnet attached to the test tube. This improvement in diagnostic methodology will strengthen efforts to control schistosomiasis.

Surface carbohydrate composition of a tapeworm in its consecutive intermediate hosts: Individual variation and fitness consequences

Carbohydrates on parasite surfaces have been shown to play an important role in host-parasite coevolution, mediating host non-self recognition and parasite camouflage. Parasites that switch hosts can change their surface molecules to remain undetected by the diverse immune systems of their different hosts. However, the question of individual variation in surface sugar composition and its relation to infectivity, virulence, immune evasion and growth of a parasite in its different hosts is as yet largely unexplored. We studied such fitness consequences of variation in surface sugars in a sympatric host-parasite system consisting of the cestode Schistocephalus solidus and its intermediate hosts, a copepod and the three-spined stickleback. Using lectins to analyse the sugar composition, we show that the tapeworm changes its surface according to the invertebrate or vertebrate host. Importantly, sugar composition seems to be genetically variable, as shown by differences among tapeworm sibships. These differences are related to variation in parasite fitness in its second intermediate host, i.e. infectivity and growth. Surface sugar composition may thus be a proximate correlate of the evolutionarily relevant variability in infectivity and virulence of parasites in different hosts.